中文摘要：

为了研究青藏高原地区太阳短波辐射,特别是其与大气因子、天空状况定量关系对高原辐射收支及其对东亚天气气候的作用,利用羊八井和纳木错2个高原腹地站一年的高时间分辨率观测资料并结合NCEP/DOE再分析资料,分析了太阳总辐射的变化特征,并通过短波通量透过率分析了各因素对短波辐射的作用。结果表明,羊八井和纳木错晴天通量透过率均较高,分别达0.807和0.817,且变动性很小,与两站地形高度和纬度有明显关系;云的存在使两站接收到的短波辐射的差异减小,羊八井和纳木错的年平均通量透过率分别为0.674和0.675,高原云的存在与变化使通量透过率产生强烈的时间变化。与地面观测的太阳短波辐射资料相比,在完全无云的晴天,NCEP同化太阳短波辐射资料对高原地区的模拟效果较好,但存在偏小的误差,羊八井和纳木错分别平均偏小5.74%和8.49%;考虑云天后,NCEP再分析值与实测值对比关系的离散度大,并总体偏高,在应用时需特别注意。

英文摘要：

The information on solar radiation, especially atmospheric factors and sky condition＇s quan- titative influence on solar radiation is very important for radiation budget in the Qinghai-Xizang Plateau （QXP） as well as for weather and climate in East Asia, while such observation and detailed analysis is few until now. Nam Co and Yangbajain locate in the center of QXP, the difference between total solar radiation at Yangbajain and Nam Co is of great significance to regional characteristics of shortwave in the hinterland of QXP. Using the one-year observed data of total solar radiation of highly temporal resolution and NCEP/ DOE reanalysis data, the characteristics and variation of total solar radiation, and factor influencing the solar radiation by shortwave flux transmittance are analyzed. The results show that the shortwave flux transmittance at Yangbajain and Nam Co stations are very high, about 0. 807 and 0. 817 in completely cloudless sky, its volatility is very low and derives from different altitudes and latitudes. At the same time, the cloud can lessen the disparity of shortwave radiation at two stations, the annual mean flux transmittance is about 0. 674 at Yangbajain station, while 0. 675 at Nam Co station. Whatrs more, cloud can make dramatic fluctuations to flux transmittance over time. In addition, comparing with the ground-based data, in com- pletely cloudless sky, the total solar radiation from NCEP is favorable despite the small error that is about 5.74% at Yangbajain station and 8.49% at Nam Co station. However, in cloudy day, the relationship between NCEP and observed values becomes much discrete, and NCEP radiation has a much systematic larger deviation. Therefore caution needs to be exercised when using NCEP data to analyse solar radiation over the QXP.